Distance transform

About: Distance transform is a research topic. Over the lifetime, 2886 publications have been published within this topic receiving 59481 citations.

Boundary-aware Instance Segmentation

Abstract: We address the problem of instance-level semantic segmentation, which aims at jointly detecting, segmenting and classifying every individual object in an image. In this context, existing methods typically propose candidate objects, usually as bounding boxes, and directly predict a binary mask within each such proposal. As a consequence, they cannot recover from errors in the object candidate generation process, such as too small or shifted boxes. In this paper, we introduce a novel object segment representation based on the distance transform of the object masks. We then design an object mask network (OMN) with a new residual-deconvolution architecture that infers such a representation and decodes it into the final binary object mask. This allows us to predict masks that go beyond the scope of the bounding boxes and are thus robust to inaccurate object candidates. We integrate our OMN into a Multitask Network Cascade framework, and learn the resulting boundary-aware instance segmentation (BAIS) network in an end-to-end manner. Our experiments on the PASCAL VOC 2012 and the Cityscapes datasets demonstrate the benefits of our approach, which outperforms the state-of-the-art in both object proposal generation and instance segmentation.

A continuous skeletonization method based on level sets

Abstract: A robust and efficient method in 2D and 3D for the calculation of skeletons for arbitrary objects is presented. The method is based on the calculation of the distance function with respect to the object boundary. This is combined, in a post processing step, with a new indicator to identify the skeleton, which coincides with the singularity set of the distance map. The indicator is defined as a suitable function of certain local momenta of this distance map and allows a robust and accurate computation of the distance from the skeleton set. This distance is then extended, again via the level set method, onto the whole space. Several applications in 2D and 3D are presented.

Distance features for neural network-based recognition of handwritten characters

Abstract: Features play an important role in OCR systems. In this paper, we propose two new features which are based on distance information. In the first feature (called DT, Distance Transformation), each white pixel has a distance value to the nearest black pixel. The second feature is called DDD (Directional Distance Distribution) which contains rich information encoding both the black/white and directional distance distributions. A new concept of map tiling is introduced and applied to the DDD feature to improve its discriminative power. For an objective evaluation and comparison of the proposed and conventional features, three distinct sets of characters (i.e., numerals, English capital letters, and Hangul initial sounds) have been tested using standard databases. Based on the results, three propositions can be derived to confirm the superiority of both the DDD feature and the map tilings.

Stereoscopic map-display method and navigation system using the method

Abstract: The view point-setting part calculates the position of the view point V to be set behind the current position of the car and up in the air, based on the current position calculated by the current position-measuring part and the position correcting part. The data reading part reads out map data M about map elements such as buildings, which are located around the current position. The representing method-changing part calculates the distance or level difference between the viewpoint and each of the map elements, and, for example, changes the representation shapes of ones of the map elements near the view point, that is, in the region within a predetermined distance from the view point, to detailed shapes, and the rest ones far from the view point, that is, outside the region within a predetermined distance from the view point, to simple shapes, respectively. Finally, the display processing part composes an image of each of the map element according to the displaying mode of the map element, which is sent from the representation method-changing part, and displays the image.

3D Shape Context and Distance Transform for action recognition

Abstract: We propose the use of 3D (2D+time) shape context to recognize the spatial and temporal details inherent in human actions. We represent an action in a video sequence by a 3D point cloud extracted by sampling 2D silhouettes over time. A non-uniform sampling method is introduced that gives preference to fast moving body parts using a Euclidean 3D distance transform. Actions are then classified by matching the extracted point clouds. Our proposed approach is based on a global matching and does not require specific training to learn the model. We test the approach thoroughly on two publicly available datasets and compare to several state-of-the-art methods. The achieved classification accuracy is on par with or superior to the best results reported to date.